Multimedia Teaching Learning Methodology and Result Prediction System Using Machine Learning

Abstract

With the availability of Multimedia Technology for the teaching-learning process, it becomes easier and convenient to enhance the learning of professional courses. Technical concepts conveyed with the help of Multimedia teaching methodology (such as Google classroom, Learning Management System (LMS), virtual lab, audio, video, animations, PowerPoint presentations, YouTube, and digital lab) for the course Energy Conversion-1 assists students to understand very complex and challenging concepts easily. The existing work does not incorporate Multimedia Technology for Energy Conversion-1 course to Third Year Mechanical Engineering students in semester VI. To identify which teaching methodology is better, this work proposes the analysis of the results for different academic years adopted different teaching methodologies, namely, traditional and Multimedia based. The approach uses a hand-crafted statistical t-value measure to prove that Multimedia based teaching method is much better than a conventional teaching method. Moreover, the existing work does not incorporate the comparison of ML models to predict the end semester examination (ESE) result of the students in the course Energy Conversion-1. This work also proposes the development of the ESE result prediction system using K-NN, SVM and Decision tree ML algorithms and the comparison of K-NN, SVM and Decision tree ML algorithms to predict the ESE result for the course Energy Conversion-1. SVM, K-NN and Decision Tree model predicted the ESE result with 78%, 70% and 70% accuracy, respectively. The comparison showed that SVM model predicted result with 78% accuracy, best fitting data. Keywords—Energy Conversion-1; Machine Learning; Mechanical engineering; Multimedia Technology; Third year students